Learning context and the other-race effect: Strategies for improving face recognition

Area of Science:

• Cognitive Psychology
• Social Psychology
• Neuroscience

Background:

• The Other-Race Effect (ORE) demonstrates differential accuracy in recognizing own-race versus other-race faces.
• Prior research indicates that training with multiple, variable images enhances general face recognition capabilities.

Purpose of the Study:

• To investigate the impact of learning context (contiguous vs. distributed) on own- and other-race face recognition accuracy.
• To build upon multi-image training paradigms by examining novel learning strategies.

Main Methods:

• Two experiments were conducted with East Asian and Caucasian participants.
• Participants learned own- and other-race faces under either contiguous or distributed presentation conditions.
• Image variability was manipulated: Experiment 1 used four variable images per identity, while Experiment 2 used one image repeated four times.

Main Results:

• A robust Other-Race Effect was consistently observed across both experiments.
• The effectiveness of distributed learning depended on image variability: it improved recognition with single repeated images but not with multiple variable images.
• Training with multiple variable images generally yielded better recognition performance than training with repeated single images.

Conclusions:

• Multiple-image training and distributed learning independently enhance face recognition accuracy through distinct mechanisms.
• Multiple-image training increases perceptual tolerance for facial variations when diverse images are available.
• Distributed learning strengthens the representation of differences among similar faces, particularly when learning from a single image.